The purpose of this side project is to see if any technique can perfrom better face detection than Dlib's HOG base detector for the purpose of online proctoring.

The results show that the Caffe model of OpenCV's dnn module performs best.

Everything from code to the results are explained in this article so you can refer here if you are interested in knowing about it. https://towardsdatascience.com/face-detection-models-which-to-use-and-why-d263e82c302c?source=friends_link&sk=c9e2807cf216115d7bb5a9b827bb26f8